Ok first of all could you please explain the metalc bonds for the last case (and whether they are different from the regular bonds). Actually also explain regular bonds in pdb, afaik not all files at all in pdb record bonds and it’s up to various programs to reconstruct them on the fly which they do with various levels of fidelity…

Please add this as an addendum to the doc while giving abstract figure asciis with arrows and bona fide cif blocks and referring to categories both.

Other than that, on the surface this proposal satisfies both parties and i’d like to pursue it, thank you. Our next step here would be to compare what we came up with against stephanie’s initial proposal and see what holes we can punch in our current one and whether we are still missing (cannot robustly yet frugally – per Martin’s – express ) from the encoding chat and documents therein that i’m about to attach here. Make this a spearate section .

Another sepearte section should be the cateogry/by category comparison and assessmnet of strneght of stepahnie’s older prosposal versus ours (whether there are tradeoffs that we make differently or whether we are strictly superior in expressivity/robustness).

Create a separate doc called “reconciliation_memo_eval.md”.